244. Serial Procalcitonin Measurements for Improved Prognostic Assessment of Patients Admitted with Bacterial Pneumonia
Session: Poster Abstract Session: Diagnostics: Miscellaneous
Thursday, October 8, 2015
Room: Poster Hall
Background: Procalcitonin (PCT) is a pro-hormone that rises in response to bacterial infections including pneumonia. While PCT has applications in the initial assessment of patients with pneumonia, data is inconclusive on its use for prognostication of clinical outcomes. We propose that serial, daily PCT measurements for inpatients with bacterial pneumonia can predict mortality, intensive care unit (ICU) admission, and bacteremia.

Methods: We conducted a prospective, observational cohort study of 505 inpatients diagnosed with pneumonia based on radiographic findings. PCT was measured on hospital days one through four. Demographics, medical history, objective data, and outcomes were obtained through chart review. Final diagnosis of community-acquired pneumonia (CAP) or healthcare-associated pneumonia (HCAP) was determined by two or more physicians blinded to PCT values. The primary endpoint for the study was a composite adverse outcome defined as all-cause mortality, ICU admission, and bacteremia. We calculated regression models with area under the ROC curve as a measure of discrimination. Statistical analysis was performed using STATA 12.1.

Results: Of 505 patients, 317 patients had a final diagnosis of CAP or HCAP. No significant difference was found for PCT values between CAP versus HCAP (Mann-Whitney U test, p=0.30). Among patients meeting the primary endpoint versus those who did not, there was no significant difference in age or the proportion with underlying lung disease, diabetes, heart failure, malignancy, or smoking. There was a significant association of PCT measured at day one through day four with the composite endpoint with AUCs ranging between 0.66 and 0.69 for CAP and 0.61 to 0.62 for HCAP. Addition of initial PCT to a statistical model including the pneumonia severity index (PSI) score significantly improved prognostic performance, increasing the AUC from 0.59 to 0.66 and from 0.58 to 0.63 for CAP and HCAP, respectively. Finally, the ability of PCT to predict adverse outcomes further improves when used for patients with PSI scores greater than 130 across all days one through four with AUCs ranging from 0.71 to 0.82 for CAP and 0.60 to 0.66 for HCAP.

Conclusion: Serial measurements of PCT in hospitalized patients with bacterial pneumonia are a promising tool for predicting adverse clinical outcomes, particularly in high-risk patients.

Suzanne Mccluskey, MD1, Michael Abers, MD1, Benjamin Bearnot, MD1, Debora Hoffman, MPH2, Maria Morales, BS2, Shreya Patel, MD, MPH1, Lauren Rosario, BA3, Phillip Schuetz, MD, MPH4, Blair Parry, CCRC, BA2, Ryan Callahan, RT, BS2, Victor Chiappa, MD1, Kent Lewandrowski, MD5, William Binder, MD6, Michael Filbin, MD7, Jatin Vyas, MD, PhD, FIDSA8 and Michael Mansour, M.D., Ph.D.9, (1)Department of Medicine, Massachusetts General Hospital, Boston, MA, (2)Department of Emergency Medicine, Massachusetts General Hospital, Boston, MA, (3)Division of Infectious Diseases, Massachusetts General Hospital, Boston, MA, (4)Medical University Clinic, Kantonsspital Aarau, Aarau, Switzerland, (5)Department of Pathology, Massachusetts General Hospital, Boston, MA, (6)Department of Emergency Medicine, Alpert School of Medicine, Brown University, Providence, RI, (7)Emergency Department, Massachusetts General Hospital, Boston, MA, (8)Massachsetts General Hospital, Division of Infectious Diseases, Boston, MA, (9)Massachusetts General Hospital, Division of Infectious Diseases, Boston, MA

Disclosures:

S. Mccluskey, None

M. Abers, None

B. Bearnot, None

D. Hoffman, None

M. Morales, None

S. Patel, None

L. Rosario, None

P. Schuetz, BRAHMS/Thermofisher: Consultant and Speaker's Bureau , Consulting fee and Speaker honorarium
BioMerieux: Consultant and Speaker's Bureau , Consulting fee and Speaker honorarium

B. Parry, None

R. Callahan, None

V. Chiappa, None

K. Lewandrowski, None

W. Binder, None

M. Filbin, None

J. Vyas, Biomerieux: Investigator , unrestricted research grant

M. Mansour, Biomerieux: Investigator ,

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.